Generalized-ensemble algorithms for molecular dynamics simulations

Abstract

In complex systems with many degrees of freedom such as biomolecular systems, conventional Monte Carlo and molecular dynamics simulations in canonical ensemble or isobaric–isothermal ensemble suffer from the multiple-minima problem, resulting in entrapment in states of energy local minima. A simulation in generalized ensemble performs a random walk in specified variables and overcomes this difficulty. In this article we review the generalized-ensemble algorithms. Multicanonical algorithm is described first. In this method, a random walk in potential energy space is realized and the simulation can avoid the multiple-minima problem. We then present two new generalized-ensemble algorithms, namely multioverlap algorithm and multibaric–multithermal algorithm, which are multi-variable/multi-dimensional extensions of the multicanonical algorithm. In the former method, a random walk in overlap space is realized, and in the latter that in both potential energy space and volume space is obtained. Emphasis is laid in the description of the molecular dynamics versions of these algorithms.